DSXchange

Welcome Aboard,
Reset the job and see if any message "From previous Run" pops up. Post your finds here.

Get rid of Job2 from the stream
What is your requirment exactly. If you dont have to run Job2 then why is it even in the stream?

Multiple posts wont help in more/different replies. Please delete your Duplicate post before anyone replies to it.

Would'nt that be something

I agree with you that triggers will slow down the process. But for a considerably large processing window, its a clean process. No change detection is required. The trigger takes care of that regardless of the data size. The historical load contains millions of records, but since its the first load,...

I put a footer note now. I get that all the time.

I prefer to write a trigger at the database end. This way all I need to do is insert the incoming record. If the same exists, the trigger will retire the previous one. But I am always fascinated with the bulk update theory.

Welcome Aboard See if the natural key exists. If it does, update the record with Current timestamp in end_date. Insert the new record with Current timestamp in begin_date and null in end_date. Do the same for all dimensions. This is the general way of doing it. Now if you have multiple records with ...

No corruption for doing multiple writes to a same hashed file. The corruption that your teammate encountered must be due to some other reason.

Send a few records to peek stage and see if the values show up in the log.

If your data volume is low, set the transaction size to 0. This way the data will be committed only after successful completion of the job.
Is this a server job or a px job?

Re-name the file after each run in after job sub-routine "ExecSh".

A simple search would have turned out the answers.
Post 1
Post 2

The job type was marked as server and hence I went ahead with it. O well.

Moderator: Please delete this post.

Likewise Craig. Stick to generation of running number.